计算机科学 ›› 2016, Vol. 43 ›› Issue (8): 194-198.doi: 10.11896/j.issn.1002-137X.2016.08.039
曹如胜,倪世宏,张鹏,奚显阳
CAO Ru-sheng, NI Shi-hong, ZHANG Peng and XI Xian-yang
摘要： 针对贝叶斯网络连续节点离散化后,概念知识表达存在模糊性和随机性的问题,提出一种将云模型与EM(Expectations Maximization)算法相结合的贝叶斯网络参数学习算法。首先运用启发式高斯云变换算法(Heuristic Gaussian Cloud Transformation)和云发生器将连续节点定量样本转换成定性概念,并记录下样本对所属概念的确定度,运用确定度概率转换公式将确定度转换成相应概率；随后复制扩充样本并按概率选择所属概念；样本更新后结合EM算法进行参数优化,实现贝叶斯网络的参数学习。仿真实验结果表明,通过云模型表征概念得到的参数学习结果更加符合实际情况,参数学习精度和网络推理准确性得到了提高。
| [1] 肖秦琨,高嵩.贝叶斯网络在智能信息处理中的应用[M].北京:国防工业出版社,2012 [2] Meng Guang-lei,Gong Guang-hong.Threat assessment of aerial targets based on hybrid Bayesian network[J].Systems Engineering and Electronics,2013,2(11):2398-2401(in Chinese) 孟光磊,龚光红.基于混合贝叶斯网的空域目标威胁评估方法[J].系统工程与电子技术,2013,2(11):2398-2401 [3] Fei Zhi-gen,Xu Xiao-jie,Dong Xi-yan.The MLE Algorithm on the Bayes Network for Machine Fault Diagnosis[J].Henan Scien-ce,2007,5(1):98-100(in Chinese) 费致根,徐小洁,董喜燕.用于机械故障诊断Bayes网络的MLE学习方法[J].河南科学,2007,5(1):98-100 [4] Yan Dai-wei,Gu Liang-xian,Pan Lei.A Study of New Method for Weapon System Effectiveness Evaluation Based on Bayesian Network[J].Journal of China Ordnance,2008,3(3):209-213 [5] Li Hong,Emmanuel A,Li Ping,et al.Imputation algorithm of missing values based on EM and Bayesian network[J].Compu-ter Engineering and Applications,2010,6(5):123-125(in Chinese) 李宏,阿玛尼,李平,等.基于EM和贝叶斯网络的丢失数据填充算法[J].计算机工程与应用,2010,6(5):123-125 [6] 李德毅,杜鹢.不确定性人工智能[M].北京:国防工业出版社,2005 [7] Liu Yu-chao.Study on calculating method based on cloud model[D].Beijing:Tsinghua University,2012(in Chinese) 刘玉超.基于云模型的粒计算方法研究[D].北京:清华大学,2012 [8] Zhang Yin-yan,Li Bi-cheng,Cui Jia-wei.Method of Target Th-reat Assessment Based on Cloudy Bayesian Network[J].Computer Science,2013,0(10):127-131(in Chinese) 张银燕,李弼程,崔家玮.基于云贝叶斯网络的目标威胁评估方法[J].计算机科学,2013,0(10):127-131 [9] 茆诗松,程依明,濮晓龙.概率论与数理统计[M].北京:高等教育出版社,2011 [10] 苏良军.高等数理统计[M].北京:北京大学出版社,2007 [11] Gao Jian-guo,Cui Ye-qin.A New Discretization Method for Con-tinuous Attributes Based on Information Entriopy[J].Microelectronics & Computer,2011,8(7):187-194(in Chinese) 高建国,崔业勤.基于信息熵理论的连续属性离散化方法[J].微电子学与计算机,2011,8(7):187-194 [12] Zheng Zhong-zhi,Duan Peng,He Xuan-jun,et al.Improvement of Bayesian classification based on the cloud model theory[J].Journal of Yunnan University of Nationalities(Natural Sciences Edition),2013,2(3):226-229(in Chinese) 郑钟志,段鹏,贺宣军,等.基于云模型理论改进的贝叶斯分类算法[J].云南民族大学学报(自然科学版),2013,2(3):226-229 [13] Zhou Xuan,Wang Lei,Zhu Yan-guang,et al.A Discretization Method of Continuous Variable in Bayesian Network Parameter Learning[J].Computer Simulation,2012,6(10):136-139(in Chinese) 周旋,王磊,朱延广,等.贝叶斯网络参数学习中连续变量离散化方法研究[J].计算机仿真,2012,6(10):136-139 |
| No related articles found! |
|
||
首页